Over his roughly 20-year tenure atop JPMorgan Chase, the mistakes that still trouble Jamie Dimon are not failed deals or bad calls. They are the delays, moments when he waited too long to cut through bureaucracy or to recognize that the wrong people were in the wrong roles. In an era defined by artificial intelligence and speed, he suggests, inertia has become an unforgivable sin.
That sensibility now shapes how Dimon is positioning the largest U.S. bank for what he sees as the most consequential technological shift of his lifetime, he said Wednesday at the World Economic Forum in Davos.
The longtime bank chief said he does not view AI as a side project confined to tech but as a core tool used across the firm, shaping nearly every part of the bank’s work, from finance and human resources to risk management, marketing, and client service. Employees across JPMorgan are expected to demonstrate how AI fits into their roles, whether that means writing code, reviewing documents, supporting customers, or operating within tightly regulated systems.
JPMorgan has developed roughly 500 AI use cases and runs an internal large language model that about 50,000 employees use each week, powered by the firm’s own data, Dimon said. Many companies, he added, underestimate how quickly AI is advancing and how broadly it will reshape operations. At JPMorgan, AI is used for fraud detection, credit decisions, hedging strategies, error reduction, marketing optimization, and idea generation, with agents on the horizon that could compress decision cycles and change how clients interact with the bank’s systems.
That breadth reflects a deeper strategic anxiety. The competitive set for a global bank no longer consists only of peers like Wells Fargo or Bank of America. It now includes Stripe, PayPal, Chime, SoFi, Revolut, and a long tail of fintechs that can target slivers of a franchise or attempt to take entire businesses end-to-end.
The capital and talent flowing into AI, Dimon said, mean incumbent financial institutions can no longer rely on size or brand alone for protection. Firms that move too slowly risk losing business to faster, more focused competitors, with AI shortening the time to respond.
For all the focus on competitive advantage, Dimon was also clear about the fallout. AI will eliminate some jobs, reshape others, and create new ones. Whether people welcome it or resist it does not matter, he argued. Companies and countries will deploy the technology regardless. The risk is not that AI advances, but that it moves faster than society can adjust to the changes it brings. If technological displacement arrives in sudden, concentrated waves, Dimon warned, the consequences could be destabilizing. “You’ll have civil unrest,” he said.
To avoid that outcome, Dimon raised the possibility of phasing in the deployment of AI-driven automation by having governments work with companies to slow large-scale job losses. That could include pressure or limits on mass layoffs, alongside incentives for retraining, income assistance, and relocation if displacement accelerates too quickly. Past trade adjustment assistance efforts fell short, Dimon acknowledged, but he argued that failure does not remove the need to try again with something that actually works. He added that those decisions would likely be made most effectively at the local level, through negotiations between governments and employers, rather than through sweeping federal mandates.
Pressed on whether he would accept the government telling companies like his not to lay off large numbers of workers, Dimon said firms would agree if the alternative were social breakdown. “We would agree if we have to do that to save society,” he said. AI will not be stopped, he added. “You’re not going to slow it down.” The question, Dimon said, is whether plans are in place to manage the damage if the technology does “something terrible.”
